A Genetic Screen to Identify Mammalian Chromatin Modifiers In Vivo.

Transcription

1 Gus Frangou, Stephanie Palmer & Mark Groudine Fred Hutchinson Cancer Research Center, Seattle- USA A Genetic Screen to Identify Mammalian Chromatin Modifiers In Vivo. During mammalian development and differentiation subsets of genes are activated in a temporally or spatially restricted manner. Once established, these programs of differential gene regulation need to be maintained over many cell divisions, a process termed cellular memory. Chromatin components such as histones, are able to generate an epigenetic code that influences the processing of the underlying genetic information. The maintenance of heritable transcription states is essential for the development of all multicellular organisms, and accumulating evidence indicates that misregulation of such processes are associated with cancer and neoplasias. Some enzymes and factors that bring about these modifications have been identified; these include DNA methyltransferases, histone methyltransferases, acetylases, and deacetylases. However, to what extent epigenetic inheritance is carried and stored at the level of these epigenetic modifications is unknown. Moreover, the functional consequences of these modifications with regards to establishment and/ or maintenance of transcription states through mitosis remain elusive. Genetic approaches represent the most direct way to elucidate complex biological processes. Indeed, genetic studies have excelled in identifying factors involved in gene regulation in vivo, as clearly illustrated by a wealth of examples from the classic model genetic organisms, S. cerevisae, Drosophila and C. elegans. Particularly illustrative are genetic studies identifying the trithorax and polycomb group genes, or modifiers of position effect variegation (PEV), as transcriptional regulatory proteins that maintain body segment identity in Drosophila. However, despite these successes, the classic model genetic systems are limited in the extent to which these systems can be utilized to fully understand epigenetic regulation in mammals. First, clear differences exist between lower and higher order eukaryotes with respect to epigenetic mechanisms regulating gene expression. For example, heterochromatin has never been identified cytologically in yeast. Secondly, Drosophila and Saccharomyces do not require DNA methylation during development and differentiation, whereas targeted mutation of the DNA methyltransferase gene (DNMT1) in the mouse results in embryonic lethality. Thirdly, genome sequence data clearly illustrates that many genes involved in mammalian development have no obvious orthologs in model organisms and vice-versa. Thus, despite highly informative studies in the classic genetic systems, similar genetic approaches in mammalian systems have not been thoroughly pursued. Unfortunately, the application of molecular genetics, with the exception of mutagenesis to cultured (diploid) mammalian cells, has proven problematic and unsuccessful. The diploid nature of mammalian cells necessitates the use of non-

2 conventional approaches for the creation of recessive mutants. RNA interference (RNAi) refers to sequence specific inhibition of gene expression mediated by doublestranded RNA (dsrna) that is homologous in sequence to the inhibited gene. Although RNAi has been successfully utilized to inhibit specific genes in Drosophila, plants and C- elegans, potent and specific RNAi has not been detected in mammalian cells until recently. Duplexes of 21 and 27-nucleotide RNA s are now broadly utilized to specifically and efficiently suppress the expression of endogenous and heterologous genes in different mammalian cell lines. Recently, whole genome RNAi screens have been performed in both human and mouse cell lines. The basic tenet of the approach is to use RNAi to alter gene expression, thereby creating an identifiable phenotype in cultured cells. Retroviral systems are ideally suited for this purpose because they allow the efficient titering and transduction of complex libraries into a broad variety of target cells. By applying appropriate genetic selections, one can recover either individual genes or complex sub-libraries that have been enriched on the basis of a specific biological characteristic from a background of irrelevant clones. The proprietary FIV RNAi libraries and recovery protocols developed by SBI, allows cultured mammalian cells to be utilized as genetic organisms similar to studies in S. cerevisae and S. pombe. The overall aim of our research is to use a novel mammalian cell system (developed by Gus Frangou, FHCRC), in which heritably silenced loci are utilized as a novel biological platform for performing somatic genetic screens to identify gene products that can modulate cellular memory The strategy outlined in Figure 1d, provides a novel paradigm for identifying the factors involved in the heritable propagation of transcriptionally silent states in mammalian cells. As illustrated in Figure 1e, thirteen novel canditate genes were identified in such a screen.

3 Figure 1a pgf10a transgene cassette FRT UAS G Promoter egfpnr IRES Neomycin SD pgf10a integrated into MEF chromatin RNA splicing

4 Figure 1b ON OFF GFP + G418 R FACS sort Trans G148-selected GFP-IRES-NEO strain Silenced clone ON OFF GFP + G418 R FACS sort Cis Gal4-cDNA G148-selected GFP-IRES-NEO strain Silenced clone

5 Figure 1c (-) G418 (+) G418

6 Figure 1d GFP + G418 R Terminate NEO selection FACS sort G148-selected GFP-IRES-NEO strain Silenced GFP-IRES-NEO clone G418 selection Silenced GFP-IRES-NEO clone CELL DEATH FIV 40k Mouse RNAi Library Trans shrnai GFP + G418 R Silenced GFP-IRES-NEO clone Re-activated clone

7 Figure 1e I II III I. Clone 32 (silenced) control (no G418 selection) II. Clone 32 (silenced) + FIV sirna control + G418 selection III. Clone 32 (silenced) + FIV 40K sirna library + G418 selection I II III

8 Figure 1a Transgene expression is known to be influenced by it s genomic integration site, copy number and orientation, effects expected to confound an assessment of interactions between promoter transcription units. To eliminate these problems we have utilized a site specific integration method (RMCE), that allows the exchange of a single pre-integrated trap cassette with an alternative cassette in the same genomic position. Accordingly, we have developed a novel RMCE Poly-A trap transgene (pgf10a, Figure 1a), containing five UAG binding sites adjacent to a strong viral promoter driving the expression of a fusion between enhanced green fluorescent protein (egfp, Clontech, Palo Alto, CA) and Escherichia coli Nitroreductase (GFPNR) and the Neomycin resistance gene. GFPNR was chosen because it can easily be detected by flow cytometry and owing to its long half-life (>24 hours), allows detection of transient transcription mrna level changes without prior knowledge of the response time course. Use of the GFPNR selectable marker facilitates live, single cell analysis and activity-dependent sorting by flow cytometry (FACS). Furthermore, by combining different thresholds of fluorescence cell sorting with variable stringency in the NR selection, it is possible to identify all ranges of promoter activity, including activated, repressed and basal activation. The transgene pgf10a incorporates several additional noteworthy features. Firstly, in an attempt to increase the frequency of intragenic integration sites of the pgf10a transgene, and to enrich for unique single integrants (including both activated and silenced loci), the transgene employed in our studies consists of a GFPNR expression unit without it s own poly-a addition signal sequence. GFPNR expression is therefore stabilized only when the pgf10a transgene captures a cellular poly-a signal (Figure 1a bottom panel). Secondly, 3 RACE analysis of the GFPNR module and genomic sequencing, facilitates the identification of the integration site of the poly-a transgene. Figures 1b & c The pgf10a cassette was introduced into the chromatin of primary mouse embryo fibroblasts by RMCE. Cells were grown under G418 selective pressure, and unique single copy integrants were identified and characterized by Southern blot. Upon removal from selective pressure, a sub-population of GFP negative cells were isolated after several weeks growth (approximately 20 population doublings), and enriched by FACS. Figures 1d & e A heritably silenced clone (GF32) was transduced with a 40K Mouse FIV RNAi library obtained from SBI (MOI 1). Clones that reactivated the silenced transgene were phenotypically identified and the integrated provirus recovered, sequenced and verified. Thirteen unique clones have been isolated including the previously characterized M31 (HP1α protein).